I'm not sure if you are being serious or joking, but yes, this does turn out to be exactly the problem. The study only has data points for Denmark in 1987 and 2001, and the whole rest of the curve is extrapolated from the data from these two years.
France, the other extreme outlier and the focus of the Economist article, is extrapolated from 3 years. The original graph at least tries to show error bars, but the Economist's "cartoon" helpfully omitted them. Theoretically, there might be something useful hidden somewhere else in the paper, but the graph that the Economist focuses on is rubbish.
It really makes the data hard to believe. ~5 minutes per day of child care in 1965? Really can't conclude anything from this until we understand that number.
